Atmanity says Atmee v2 was developed over two years after Google acquired the founder's AR startup, aiming to make face-to-face the next AI interface. The company claims the model creates characters from one image plus 10 seconds of audio, with contextual expressions and gestures. In a comparison with Seedance 2.5 using the same image and prompt, Atmee v2 reportedly generated video in 3.5 seconds for about a cent versus 233.6 seconds and $0.50. Demos include a talk show, tutoring, gaming, and a live Taipei alley scene the author calls a step toward an "Agentic World Model."

AI summary of the linked source — not editorial content.

Explore the topic

Loading related coverage…

Comments

Loading comments…